Leveling the playing field - the economics of electricity generation in Europe


Published on

A calculation of the costs of electricity generation in Europe, including externalities

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Leveling the playing field - the economics of electricity generation in Europe

  1. 1. Leveling the playing field - An economic assessment of electricitygeneration in EuropeDr David Clubb, CPhys, FIWA ndPublished 22 August, 20121. AbstractThe calculation of the externalities of electricity production is of great importance in shaping the views ofbusinesses, policy-makers and the public in relation to the value of energy. This paper attempts to draw togethera number of existing studies on aspects of electricity generation in the European Union in order to calculate an‘Extended Levelised Cost of Electricity’ (eLCOE) which takes account of a wide range of externalities. I usebenefit transfer to calculate a European average value for a number of generating technologies, bothconventional and renewable. I find that the external costs are a significant component of overall electricityproduction costs, particularly for conventional generation, but also for biomass. I also estimate the employmentvalue of various electricity generating technologies, and estimate overall costs of electricity in 2020 bytechnology.The average European levelised cost of electricity by generating type, including externalities, is shown in thefigure below. Onshore wind provides the least-cost electricity generation, and also rates highly in employmentper unit electricity generation and cost. eLCOE in 2010 (€/GWh, 2005 prices) External 450,000 400,000 impact 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 -50,000Extended levelised cost of electricity (eLCOE), i.e. the LCOE plus externalities. Negative externalities indicatethat there is an economic benefit above and beyond the production of electricity itself. The error bars indicatethe range of costs arising from a sensitivity analysis. See full text for details.
  2. 2. Employment per unit electricity generation and cost. See full text for details2. IntroductionEurope’s energy system is entering a transition phase. The established mechanisms of centralised generation,and top-down transmission and distribution will see a shift unprecedented in the recent history of energy use.Distributed and intermittent renewable generation will become increasingly important components of theelectricity mix, contributing to energy security and reducing the carbon content of electricity.However, the field in which renewable energy operates is still not level with respect to its competitors.Conventional electricity generators benefit economically from an externalisation of impacts which are notperceived as direct costs to the generator. When costs are not properly allocated to generators, the economicsof the electricity sector are skewed against renewable energy generators, which – by and large – have far lowerexternal costs than traditional electricity generators.Economic aspects are important, because decisions which are made on the basis of incomplete information canresult in perverse outcomes. For example, the fact that environmental impacts are generally not included in thecosts of coal extraction and combustion means that coal-fired power stations are more likely to be installed,because the final product – electricity at the point of use – contains an implicit subsidy which is not borne by theuser or the generator. If better information is made available, then decisions are also likely to be improved.This paper aims to help make the move towards more complete information on external costs by summing theoutputs of existing studies, or ascribing a value to externalities which are not generally accounted for. It is ahighly complex topic, and I am attempting no more than to draw together the most up to date literature in orderto help inform current thinking.3. MethodologyUnitsThe Levelised Cost of Electricity (LCOE) is the usual way to compare the cost of generating electricity fromdifferent sources. It represents the present value of the total cost of building and operating a generating plantover an assumed financial lifecycle, converted to equal annual payments and expressed in terms of Euros at afixed point in time (herein 2005 unless otherwise stated). I adopt this standard, although I also derive anadditional form – the Extended Levelised Cost of Electricity (eLCOE) – in order to compare differenttechnologies. The eLCOE includes many of the calculable externalities which arise from electricity generation. Iuse GWh as the baseline energy unit, and derive all other outcomes from this. This approach allows maximumutility in extrapolating these figures for use in different Member States or geographical regions.Geographic boundariesI take the European Union Member States as the boundary in terms of data on impacts, although where justifiedI use benefit transfer from other regions such as the USA. The use of average figures for the EU is a necessary 1
  3. 3. over-simplification, even where some impacts, such as employment, load factors of generating capacity, healthimpacts from air pollution, and water use, are highly location-specific. The methodology represents an approachwhich could be adopted at a local or regional level in order to derive more appropriate assessments, particularlyconsidering the geographically-linked nature of many energy sources.External impact and electricity generation typesThe external impacts considered in this paper are Balancing Costs, CO2 Cost, Damage to Materials, Fuel CostVolatility, Merit Order Effect, Morbidity and Mortality, Risk Underwriting Cost, Security of Supply, Virtual Grid, andWater Use Cost.Balance of Payments and Ecosystem Impacts were also considered. They have not been included, as theeconomic impact of a negative balance of payments is unclear, and the economic impacts of ecosystemdamage from electricity generation is incalculable with present knowledge.The technologies considered are Biomass, Coal (pulverized coal combustion), Geothermal, Natural Gas(combined cycle), Nuclear, Hydro (small and large), Solar PV and Wind (onshore and offshore). To simplify theoperation, I take the ‘best practice’ within each sector where there are multiple technology types.I do not consider Concentrating Solar Power, as it is largely predicted to be an ‘import’ technology (i.e. much ofthe electricity consumed within Europe will be imported from Turkey or further afield), and deployment will be 1low by 2020 .Cost of electricityA number of organisations provide figures for the LCOE, both for the existing ‘state of the art’ and projectingcosts to 2020. These costs are summarised in Figure 1 and Figure 2.Figure 1: Comparison of LCOE calculated by different Figure 2: Predicted LCOE in 2020; sources as in 2 3 4 5 6 7 8studies (EC , IEA , CCC , BNEF , ECN , EPIA , JRC ) 9 Figure 1 except for ECF and EGEC 10For the purposes of calculating the overall cost of electricity I take a numerical average for each technology(Table 1).1 International Energy Agency, ‘Technology Roadmap - Concentrating Solar Power’, 2010.2 European Commission, ‘Energy Sources, Production Costs and Performance of Technologies for Power Generation,Heating and Transport’, 2008.3 International Energy Agency, ‘Projected Costs of Generating Electricity’, 2010.4 Committee on Climate Change, ‘Costs of Low-carbon Generation Technologies’, 2011.5 M Liebreich, ‘Bloomberg New Energy Finance Summit; Keynote Speech’, 2011.6 European Climate Foundation, ‘A Zero-carbon European Power System in 2050: Proposals for a Policy Package’, 2010.7 European Photovoltaic Industry Association, ‘Solar Photovoltaics; Competing in the Energy Sector’, 2011.8 R Arantegui, ‘Current Costs of the Geothermal Power Technologies’ (presented at the Geothermal electricity workshop,Brussels, 2011).9 European Climate Foundation, ‘A Zero-carbon European Power System in 2050: Proposals for a Policy Package’.10 Jean-Philippe Gibaud, ‘European Vision for Geothermal Electricity Development’ (presented at the Geothermal electricityworkshop, Brussels, 2011). 2
  4. 4. Table 1: LCOE values for differenttechnologies (2005 prices)Energy LCOE/GWh LCOE/GWhtechnology (2007) (2020)Biomass 136 500 142 500Coal 87 500 94 333Geothermal 112 677 97 500Hydro (large) 90 000 85 000Hydro (small) 95 500 83 500Natural gas 85 000 103 333Nuclear 79 667 83 333Solar PV 419 213 218 750Wind (offshore) 100 000 95 000Wind (onshore) 82 600 66 750The fact that renewable technologies see such a high relative drop in their levelised cost by 2020 should notcome as a surprise; the learning rates for these technologies are acknowledged to be much higher than their 11 12 13conventional counterparts , , . The cost of biomass electricity is an exception, and this may reflect itssensitivity to fuel prices.Sensitivity analysisEach of the variables within the total cost has an uncertainty associated with it. The total external cost, andtherefore eLCOE, has a value which is sensitive to each of the inputs. A sensitivity analysis studies how thevariation of each of the variables affects the overall output.To take an example, the cost of CO2 has been calculated as anything from €0 to over €1 500 per metric tonne. Asensitivity analysis would recalculate the impacts of electricity generation using various possible values for eachparameter, and demonstrate a range of outcomes.A sensitivity analysis with two scenarios, ‘renewable friendly’ and ‘conventional friendly’, has been carried out onthe data in this paper. The assumptions are provided in Table 2.Table 2: Sensitivity analysis assumptionsVariable Renewable Conventional Comment scenario scenarioBalancing Half Double Balancing costs are not easily quantified; all generatorscosts need to be balanced against each other, including conventional fuels. This simple approach suggests two regimes which represent higher or lower costs than that selected from the literature.CO2 impact €68/ tonne €12/ tonne Low value represents current CO2 market values, which are likely to underestimate the full consequences of CO 2 emission. High value is the Euro equivalent of the value in the Stern Report ($85/ tonne)Pollution Double Half The renewable scenario assumes that the environmental damage arising from pollution doubles the human health impact; conventional scenario assumes that the health impacts are overstatedHedging Double Half The renewable scenario takes into account some aspect of the balance of payments impact (i.e. the additional cost of servicing imports); conventional value assumes a scenario of much more stable future fossil fuel pricesOccupational No change Marginal impact to overall costsmortalityMaterials No change Marginal impact to overall costsdamage11 H. Winkler, A. Hughes, and M. Haw, ‘Technology Learning for Renewable Energy: Implications for South Africa’s Long-term Mitigation Scenarios’, Energy Policy, 37 (2009).12 European Photovoltaic Industry Association, ‘Solar Photovoltaic Electricity: A Mainstream Power Source in Europe by2020’, 2009.13 United Nations Environment Programme, ‘Renewable Energy: Investing in Energy and Resource Efficiency’, 2011. 3
  5. 5. Merit order Double Half Renewable scenario takes a less conservative assessment, and broadens the value to all electricity rather than weighted by renewable installation level (see comment on page 9). Conventional scenario assumes that the benefits of the Merit Order effect are exaggeratedRisk Double Half Renewable scenario assumes a higher level of risk forunderwriting nuclear electricity than calculated, or could also be considered as a proxy for inclusion of some decommissioning costs. Conventional scenario assumes that risks are overstatedVirtual grid No change Marginal impact to overall costs 3 3Water use €1.2/ m €0.05/ m The renewable scenario uses the Italian ‘water stress’ 14 scenario ; low value uses OECD averagePrevious studiesPrevious studies have considered electricity production externalities (see Table 3). However, they generallyconsider the situation in one Member State, or do not consider wider factors such as water use in the course ofelectricity generation.Table 3: Summary of existing studies on electricity externalitiesReport Area Comment 15ExternE (2005) Europe Calculates externalities for specific impacts (such as air pollution, nuclear accident, noise etc.), but does not provide an overall figure for different technologiesUmweltbundesamt Germany Germany only; includes a summary of six previous externality reports; 16(2007) 3% discount rate (0-20 years), 1.5% discount rate (>20 years), 0% discount rate (inter-generational). Compares fossil fuel and renewable technologies 17RECABS (2007) Europe Online calculator which can be used to determine externalities according to different variables. Documentation provides reference scenarioNational Research USA USA-based study which cannot be compared with European figures due 18Council (2009) to very different environmental and economic factorsDanmarks Denmark Denmark only; focus on airborne pollutantsMiljøundersøgelser 19(2010)Vlaamse Flanders Flanders onlyMilieumaatschappij 20(2011)4. Results a. Balancing costsA secure and reliable electricity grid requires that demand and supply match on an instantaneous basis. Thetraditional method of operation was straightforward, in that large generators were able to match the changingload on the network because they had well-defined outputs which could precisely match the predicted demand.This system is changing as a result of the changing profile of electricity generators. Variable and intermittentgenerators, particularly wind and solar, are more difficult to fit to the demand because the difference betweenforecasted generation and actual supply is much more likely to be significant than for traditional generators. Thisgives rise to an additional need for balancing capacity. However, the potential gap between supply and demand14 E.M. Jenicek et al., ‘Army Overseas Water Sustainability Study’, US Army Corps of Engineers (2011).15 ExternE, ‘Externalities of Energy’, 2005.16 Federal Environment Agency, Germany, ‘Economic Valuation of Environmental Damage; Methodological Convention forEstimates of Environmental Externalities’, 2007.17 RECABS, ‘Interactive Renewable Energy Calculator - REcalculator - Renewable Energy Cost and Benefit to Society’, n.d.18 National research council of the National Academies (USA), ‘Hidden Costs of Energy: Unpriced Consequences of EnergyProduction and Use’, 2009.19 Danish Environment Ministry, ‘Environmental costs of emissions’, 2010.20 Vlaamse Milieumaatschappij, ‘Damage costs of current and future electricity production in Flanders’, 2011. 4
  6. 6. is likely to be mitigated by an increasing sophistication in energy use, through consumer awareness coupledwith smart metering systems. Improved prediction of output will also mitigate this impact.The amount of balance capacity increases as the amount of renewable energy generation increases, althoughthe relationship is not strictly linear. A simplified model of Europe with 100% renewable electricity, supplied by 21wind and solar power, suggests that the storage energy capacity required would be up to 15% of demand . Thestorage (or balancing power) required depends on the energy mix, because some generation types arecomplementary, with a decrease in one generator often being compensated for by an increase in another (seeFigure 3). It is also likely that demand-side management, including measures such as smart metering whichallow real-time alteration of electricity use in response to price signals, will improve the ability of the electricitysystem to adjust to fluctuating input, and hence reduce the requirement to add balancing capacity. 22Figure 3: Monthly capacity factor for wind and solar in GermanyThe apportioning of fees for balancing capacity varies by Member State. In some countries the fees are passedon to all consumers by the TSO; in this case, there is an implicit subsidy to the generators which require the 23biggest balancing capacity. The method of apportioning costs varies greatly between Member States , whichmeans that providing exact numbers for each technology is not possible.I apply the cost of €4 000/GWh which is suggested by the upper end of estimates from the European Wind 24Energy Association (Figure 4), and a more modest value for solar PV estimated as €1 200/GWh .21 D. Heide et al., ‘Seasonal Optimal Mix of Wind and Solar Power in a Future, Highly Renewable Europe’, RenewableEnergy, 35 (2010).22 European Photovoltaic Industry Association, ‘Set for 2020’.23 ENTSO-E, ‘Overview of Transmission Tariffs in Europe: Synthesis 2010’, 2010.24 Landesbank Baden-Wurttemberg, ‘Photovoltaics Sector: Valuing the Invaluable’, 2008. 5
  7. 7. 25 Figure 4: Result of meta-study on balancing costs for a number of wind regimes b. CO2 The economic impact of anthropogenic CO2 emissions has been placed within a wide spectrum, from zero to many hundreds of US dollars per tonne. Discussions surrounding the impact are likewise characterised by wide variety and I do not intend to reproduce them here. I follow the approach adopted in the EEA publication “Costs 26 of Air Pollution from Industrial Facilities in Europe” which uses the short term traded price of carbon in the UK in 2020, set at €33.6/tonne. The EU Emission Trading Scheme (ETS) is currently operating in its second Trading Period, which runs from 2008 until the end of 2012. The vast majority of permits in this Trading Period were freely allocated to participants in the ETS, so although there is a ‘price’ for carbon, it is not currently borne by the producer, and can be considered as an externality. I therefore make no distinction between the monetised costs of CO2 produced by ETS participants (generally coal, natural gas and co-fired biomass) and that from lifetime emissions from renewable electricity generators. Knowing the Life Cycle Emissions (LCE) of different electricity generating technologies, I can calculate the cost of CO2 emissions for each type of generator (Table 4). 27 Table 4: LCE of different generating technologies; from average values within Kenny et al , except where otherwise stated 28 31 32Energy Biomass Coal Geothermal Hydro Hydro Natural Nuclear Solar Wind Wind 29 30technology (large) (small) gas PV (offshore) (onshore)LCE 32 925 15 32 11 437 82 40 30 11(tCO2/GWh)CO2 emission 1 075 31 080 504 1 075 370 14 683 2 755 1 344 1 008 370cost (€/GWh) c. Damage to materials In 2008, AEA Technology carried out a modeling exercise to evaluate the impacts of PM 2.5, SO2, NOx, VOC and 33 NH3 on human health and materials . These are the pollutants covered by the Thematic Strategy on Air 25 European Wind Energy Association, ‘Powering Europe: Wind Energy and the Electricity Grid’, 2010. 26 European Environment Agency, ‘Revealing the Costs of Air Pollution from Industrial Facilities in Europe’, 2011. 27 R. Kenny, C. Law, and J.M. Pearce, ‘Towards Real Energy Economics: Energy Policy Driven by Life-cycle Carbon Emission’, Energy Policy, 38 (2010). 28 Does not include Indirect Land Use Change emissions 29 Hanne Lerche Raadal et al., ‘Life Cycle Greenhouse Gas (GHG) Emissions from the Generation of Wind and Hydro Power’, Renewable and Sustainable Energy Reviews, 15 (2011). 30 Compatible with A.F. Sherwani, J.A. Usmani, and Varun, ‘Life Cycle Assessment of Solar PV Based Electricity Generation Systems: A Review’, Renewable and Sustainable Energy Reviews, 14 (2010). 31 Hermann-Josef Wagner et al., ‘Life Cycle Assessment of the Offshore Wind Farm Alpha Ventus’, Energy, 36 (2011). 32 Raadal et al., ‘Life Cycle Greenhouse Gas (GHG) Emissions from the Generation of Wind and Hydro Power’. 33 S Pye et al., ‘Analysis of the Costs and Benefits of Proposed Revisions to the National Emission Ceilings Directive. NEC 6
  8. 8. 34Pollution (TSAP), and the power sector is responsible for a significant proportion of them. Damage to crops isalready considered under the CAFE modeling programme, and is therefore included within the ‘Morbidity andMortality’ section (Page 9). 35The estimated annual damage to materials from SO 2 is €1.1 bn in 2000, and €0.7 bn in 2020 . I assume alinear change in the SO2 emissions giving an annual cost of €0.94 bn for 2008. The sectoral emissions for the 36same year totalled 3 144 000 tonnes , which means that each emitted tonne of SO 2 had an economic cost onmaterials, due to the secondary impacts of acid damage, of €299.Biomass, coal and gas-fired electricity account for the electricity sector SO 2 emissions, with emission 37coefficients of 11, 310 and 0.3g/GJ respectively .The 2009 EU-27 output of heat from combustible fuels was 405 166 GWh, and from electricity was 456 873GWh. It is not possible to disaggregate the inputs to the heat and electricity sectors, so I make a simpleapportion of damage according to the output. I therefore calculate that 53% of the damage from total SO2emissions comes from the electricity generation sector, which underestimates its contribution, as efficiency incombustion to generate electricity is generally lower than efficiency of heat production.This provides the final damage costs of €6/GWh for biomass, and €157/GWh for coal. d. Fuel cost volatility 38A 10% increase in the price of oil is calculated to decrease European GDP by 0.5% ; given a number of recentoil price shocks, there are clear economic benefits to be obtained from diversifying from those energy sourceswhich are most heavily dependent on this sector. In the context of this study, the impact acts predominantly onnatural gas-fired electricity, the value of which is strongly coupled to the price of oil. To a lesser extent it impactson the other ‘fuel-using’ technologies which use significant quantities of petroleum products in the fuel supplychain, namely biomass, coal and nuclear.The hedging value of renewables is more subtle than just lessening exposure to volatility or high fuel prices. 39High oil prices are linked with a significant economic impact in most oil-importing countries , and theperformance of renewable electricity generation can therefore be seen as having a strongly counter-cyclical 40impact which further augments its benefits. The hedging benefits may have been first described by Lind , whodescribed renewable energy investments as a form of ‘national insurance’. Wider macro-economic benefits froma diversification from fossil fuel include mitigating inflation and interest rate increases, and reducing shockimpacts to stock markets. The economic impact of oil price changes is discussed in detail by Awerbuch and 41Sauter .The same authors calculate that the offsetting of oil-induced macro-economic losses are worth $250-$450/kWhfor renewable electricity, which is subdivided into wind and solar ($200/kWh), and geothermal and biomass($800/kWh) on the basis of capacity factors (23% and 92% respectively). Nuclear electricity is included withinthe $800/kWh ‘offset’ section. 42The RECABS assessment converts this ‘installed capacity’ value to a kWh rate assuming a 5% discount rateand 20 year life-span, and incorporates it as an additional externality to the cost of gas and coal-fired electricitygeneration, at rates of €7 000/GWh and €2 300/GWh respectively. e. Merit order effectSeveral European studies have concluded that renewable energy installations with no fuel costs, such as wind 43,44energy, act to suppress prices for the consumer . This is because they are used preferentially to supplyCBA Report 3’ (DG Environment, 2008).34 European Commission, ‘Thematic Strategy on Air Pollution’, 2005.35 P. Watkiss, S. Pye, and M. Holland, ‘CAFE CBA: Baseline Analysis 2000 to 2020’ (CAFE programme, 2005).36 European Environment Agency (Copenhagen)., ‘Air Pollutant Emissions Data Viewer (NEC Directive)’, 2010.37 European Environment Agency (Copenhagen)., Combustion in Energy Transformation Industries; Guidebook, 2009.38 International Energy Agency, ‘Analysis of the Impact of High Oil Prices on the Global Economy’, 2004.39 R. Jiménez-Rodríguez, M. Sánchez, and European Central Bank, Oil Price Shocks and Real GDP Growth; EmpiricalEvidence for Some OECD Countries (European Central Bank, 2004).40 R Lind, Discounting for Time and Risks in Energy Policy (Washington D.C.: Resources for the future, 1982).41 S Awerbuch and R Sauter, Exploiting the Oil-GDP Effect to Support Renewables Deployment (University of Sussex,SPRU - Science and Technology Policy Research, 2005).42 International Energy Agency, ‘Renewable Energy Costs and Benefits for Society - RECABS’, 2008.43 W. Steggals, R. Gross, and P. Heptonstall, ‘Winds of Change: How High Wind Penetrations Will Affect Investment 7
  9. 9. electricity when they are generating, which tends to reduce the exposure of consumers to the marginalgenerators, such as open-cycle gas turbines, which are very costly. This effect is known as the ‘merit ordereffect’, because it describes the way in which the merit (or cost) of a generator is used to define the stage in theload profile when it should become operational. An example of this is seen in Figure 5. Wind and hydro, whichhave near-zero marginal costs, come first in the merit order. Nuclear electricity is generally used as baseload,due to its inability to significantly modulate its output (both for economic and practical reasons) and because thefuel cost is a very small proportion of the overall energy price. Thus, nuclear electricity also comes early in themerit order. 45Figure 5: Merit order of marginal production at the Nordic Nord Pool MarketThe case of wind energy has been studied far more than any other generator. In this case, the magnitude of thebenefit is about 1% of the spot value of electricity for each percentage point of installed wind capacity (wind isgenerally the subject of these studies as it has significant penetration in several Member States). The exampleof the effect of wind on electricity prices in West Denmark can be seen in Figure 6. Similar conclusions can be 46drawn about the case of wind power in Ireland; a 2011 paper shows that the reduction in wholesale marketprice of electricity produces a benefit to the consumer greater than the balancing costs; in other words, the 47additional benefits of reduced CO2 emissions and improved energy security come at no cost .The merit order effect can only provide some net economic benefit in relation to an average annual electricityvalue. In Figure 5 the average demand is signified by the line at 410TWh/year so anything below that lineprovides a net benefit. This includes all renewables which operate with near-zero marginal cost, such as solarphotovoltaic.Incentives in the GB Electricity Sector’, Energy Policy (2010).44 European Wind Energy Association, ‘Wind Energy and Electricity Prices; Exploring the “Merit Order Effect”’, 2010.45 H. Lund et al., Danish Wind Power-Export and Cost (Institut for samfundsudvikling og planlaegning, Aalborg Universitet,2010).46 E Clifford and M Clancy, ‘Impact of Wind Generation on Wholesale Electricity Costs in 2011’ (Sustainable Energy Authorityof Ireland, EirGrid, 2011).47 International Energy Agency, ‘Climate and Electricity Annual’, 2011. 8
  10. 10. 48Figure 6: Impact of wind power on spot electricity prices in West Denmark, December 2005I take this opportunity to clarify that, although the merit order effect applies to all electricity prices, rather thanjust those associated with each technology, I cannot logically extrapolate this value across the whole of Europe,as the installed capacity of each generator type is different in each Member State. Thus I calculate the benefitfor each generator due to its installed capacity (i.e. on a ‘per GWh’ basis). This has the effect of significantlyunderestimating the actual benefit. Table 5: Economic impact of the Merit Order Effect 50 Biomass Coal Geothermal Hydro Hydro Natural Nuclear Solar Wind 49 (large) (small) gas PV€/GWh n/a n/a n/a Unknown Unknown n/a Unknown 10,000 11,000It seems reasonable to assume that there is also a merit order effect for nuclear and large hydro; however, thereis no mention of this in the literature, and they have therefore not been included in the overall total.Note that as the renewable proportion of Europe’s electricity increases, the merit order effect lessens. This isbecause additional renewable capacity decreases the ‘full load hours’ (a measure of the annual usage ofgenerators) of the plants towards the right of the merit order (Figure 5) which decreases the incentive to installconventional plant.As the proportion of low-margin production increases, the electricity price will start to track variable costs less,and fixed costs more. This will be beneficial to consumers as price stability is locked into the system, but therelative advantage of low marginal-cost generators due to the merit order will be lost.An additional and significant impact of the growing penetration of renewable electricity will be to increase theamortisation risk for fuel-consuming generators which will escalate investment costs for new generating plant. f. Morbidity and mortalityThe choice of the value of a statistical life (VSL) significantly influences the economic impacts of mortality and 51health-related factors. The most recent OECD meta-study suggested a median value of $2,814,000 52(equivalent to €2,181,412 ). Although different studies demonstrate some variation in the value ascribed tohuman life, this does not significantly affect the relative order of the electricity generators in the economic impactassessment.The ExternE project states that pollution deaths need to use the Value of a Life Year (VOLY), because “the loss 53per air pollution death tends to be small”. The value of €40 000 for the EU-25 was obtained in a survey carriedout in nine European countries on ‘contingent valuation’. This method asks how much people would be willing to48 European Wind Energy Association, ‘Wind Energy and Electricity Prices; Exploring the “Merit Order Effect”’.49 International Energy Agency, ‘Renewable Energy Costs and Benefits for Society - RECABS’, p. 40.50 European Wind Energy Association, ‘Wind Energy and Electricity Prices; Exploring the “Merit Order Effect”’.51 H Lindhjem, S Navrud, and Nils Braathen, ‘Valuing Lives Saved from Environmental, Transport and Health Policies; aMeta-analysis of Stated Preference Studies’, 2010.52 Assuming 2% inflation 2010-2011, and an exchange rate of 0.76 in 201053 ExternE, ‘Externalities of Energy - Frequently Asked Questions’, n.d. 9
  11. 11. pay to reduce health risks associated with air pollution.In this study I use VSL where there are immediate economic impacts from lives lost (CO 2, occupationalmortality), and VOLY for the economic impacts from pollution which causes long-term reduction in lifeexpectancy.Occupational hazardThe renewable energy fuel supply chain is short or non-existent compared with fossil fuel and nuclear electricity.The expansion of the renewable energy sector thus has potentially significant occupational health benefits. Theobvious exception to this is the biomass sector, which potentially includes the occupations of farming, forestryand combustion. This accounts for its relatively high mortality rate, second only to the coal supply chain.There are cogent arguments within the literature that there is some degree of internalisation of the economicimpact of morbidity in the energy sector, through wage values, education and training. In other words,participants within this sector receive high wages (which are a recognition of the increased risk of theirprofession), and there is an additional training burden on those businesses to properly prepare their employeesfor work within a relatively hazardous environment. In line with the literature, I therefore calculate the cost of 54occupational morbidity as 20% of the VSOL value . Table 6: Economic impact of mortality and morbidity via the fuel supply chain; biomass, coal, natural gas, 55 nuclear figures taken from ExternE , and geothermal, large hydro, solar PV and wind figures taken from the 56 SECURE project Biomass Coal Geothermal Hydro Hydro Natural Nuclear Solar Wind (large) (small) gas PV -4 -4 -7 -6 -5 -5 -8 -7Mortality 1.6×10 3.1×10 2.0×10 9.7×10 Unknown 4.6×10 1.7×10 2.8×10 3.4×10/GWhCost 74 142 0 4 0 21 8 0 0€/GWhAlthough the occupational benefits are small for some sectors compared with a reduction in by-productsassociated with combustion, in terms of directly perceived benefits the fuel supply chain is highly important: “Whereas many health benefits associated with a reduction in high carbon dioxide emission energy production may be perceived by some as distant or uncertain, prevention of deaths of energy workers as a result of an improved occupational safety profile of renewable technologies has the 57 potential to be immediate, obvious, and sizable”.Societal costsThe by-products of electricity production take various forms, but they are generally detrimental to theenvironment. Therefore, human health is also affected; the pathways by which this occurs are well summarised 58in .Those pollution impacts which can be quantified are summarised in Table 7.Table 7: Mortality and morbidity levels, and economic impacts, of electricity generation (excluding supply chain). 59Economic impact taken from Economic damage, €/GWhPollutant Damage cost Biomass Coal Geothermal Hydro Natural Nuclear Solar Wind (€/t) (both) gas PV (both)SOx 6 381 253 7 121 neg. neg. 7 neg. neg. neg.NOx 6 263 4 757 3 382 neg. neg. 2 007 neg. neg. neg.NMVOC 513 13 1 514 neg. neg. 3 neg. neg. neg.PM10 13 851 1 895 997 neg. neg. 45 neg. neg. neg.PM2.5 21 331 2 534 691 neg. neg. 69 neg. neg. neg.Pb 965 000 73 28 neg. neg. 1 neg. neg. neg.54 ExternE, ‘Externalities of Energy’, p. 210.55 ExternE, ‘Externalities of Energy’.56 P Burgherr, P Eckle, and S Hirschberg, ‘Final Report on Severe Accident Risks Including Key Indicators’ (Secure project,2011).57 Steven Sumner and Peter Layde, ‘Expansion of Renewable Energy Industries and Implications for Occupational Health’,Journal of the American Medical Association, 302 (2009).58 D. Pudjianto et al., ‘Costs and Benefits of DG Connections to Grid - Studies on UK and Finnish Systems’, 2006.59 M Holland, ‘Costs of Air Pollution from Industrial Facilities in Europe’ (European Environment Agency, 2011). 10
  12. 12. Cd 29 000 0 0 neg. neg. 0 neg. neg. neg.Hg 910 000 5 5 1 neg. 0 neg. neg. neg.As 349 000 12 10 neg. neg. 0 neg. neg. neg.Cr 38 000 1 1 neg. neg. 0 neg. neg. neg.PCDD/F 27 000 000 000 4 860 972 neg. neg. 49 neg. neg. neg.Benzo(a)yprene 1 279 000 5 3 neg. neg. 3 neg. neg. neg.Indenopyrene 1 279 000 2 6 neg. neg. 4 neg. neg. neg.Radioactive 0 0 0 0 0 6 000 0 0waste cost Total 14 411 14 731 1 0 2 187 6 000 0 0Although combustion technologies are currently responsible for the majority of pollution from electricity 60generation, the trend for harmful by-products is continuing to decrease , and the future economic impact fromthese sources will follow that trend.In general, lifetime emissions for renewables except geothermal (secondary emissions) and biomass arenegligible. The most significant pollutant during production of geothermal electricity is H 2S, but no value couldbe found in the literature to calculate its economic impact.Many additional health effects are outwith the scope of this paper, due to the complexity of the systems. Theyinclude:  Ozone production (local) and depletion (global)  Secondary effects which may also indirectly affect human health, such as: o Climate change-related issues such as temperature change, extreme weather events, sea-level rise o Ecological issues, such as deforestation, desertification, biodiversity, disease and changes in agriculture practiceNuclear pollutionNuclear plants are a special case in terms of pollution, as the quantity and impact of radioactivity is more difficultto quantify and monetise. There is, however, a significant body of literature which has been produced looking atexactly this issue. 61The issue of waste was considered by the CE Delft paper , which attempted to create an average value for theimplicit subsidy to managing and storing high-level waste. The paper considered the fact that funding of currentradioactive waste management is partially or fully accounted for by the operators of nuclear facilities in somecountries, whilst final waste management costs are not. It is undeniable that nuclear waste will continue to needhandling, storage and management, and the fact that this is not being funded through current electricitygeneration costs mean that there is an implicit subsidy being levied on future storage costs. This subsidy iscalculated as €3 000/GWh.The subsidy for future waste management is exclusive of the impact of radioactive emissions from the milltailings, which have a very small impact but last for an extremely long time. These have been calculated at €3 62000/GWh . These costs may just be the tip of the iceberg; the Nuclear Energy Agency of the OECD reproduceda table in its 2003 paper on nuclear externalities demonstrating that the tailings account for a very smallcomponent of overall radioactive emissions; 97% of which arise from electricity generation and fuel 63reprocessing .Decommissioning costs, which are partly paid for under an operating tax in some countries, are not quantifiable,but almost certainly represent another subsidy by the State to nuclear power operators. The case of theDepartment for Energy and Climate Change in the UK is an illustrative one. In 2009/10, it saw 56% of its net 64operating costs allocated to the Nuclear Decommissioning Authority . In other words, the UK taxpayer is stillpaying large amounts for the legacy of activities in the civil nuclear sector more than 50 years ago; the total60 European Environment Agency, ‘European Union Emission Inventory Report 1990 — 2008 Under the UNECE Conventionon Long-range Transboundary Air Pollution (LRTAP)’, 2010.61 B. A. Leurs et al., Environmentally Harmful Support Measures in EU Members States (CE, Solutions for environment,economy and technology, 2003), p. 145.62 International Energy Agency, ‘Renewable Energy Costs and Benefits for Society - RECABS’, p. 33.63 Nuclear Energy Agency, ‘Nuclear Electricity Generation: What Are the External Costs?’, 2003.64 Department for Energy and Climate Change, ‘DECC Resource Accounts 2009-2010’ (UK Government, 2010). 11
  13. 13. 65liability for UK nuclear sites is estimated as €51bn . This is an example of mortgaging the wealth of futuregenerations against the lower cost of electricity to present consumers. g. Risk underwritingThe operators of fossil fuel and nuclear plants need to insure against far higher risks than their renewableenergy counterparts. Whilst fossil fuel and renewable energy plant operators must accept the full liability, this isnot generally the case for nuclear plant operators.Nuclear operators are bound by national policy, which in turn is generally defined by an international agreement.The international legal framework is a complex patchwork of different conventions, which makes it difficult tounderstand the liabilities of each party. However, it is clear that the variety of liability between Member Statesmeans that the amount of risk underwriting likewise varies by Member State.This variation in liability has been highlighted as a matter of concern by the European Economic and SocialCommittee: “A harmonised liability scheme, including a mechanism to ensure the availability of funds in the event of damage caused by a nuclear accident without calling on public funds, is in the view of the EESC also essential for greater acceptability of nuclear power. The current system (liability 66 insurance of $700 million) is inadequate for this purpose” 67A study carried out for the European Commission in 2003 reported a premium of €190/GWh to cover the 68limited national and international liabilities of €1 500m for France. This is equivalent to €197/GWh in 2005values. Assuming a ‘liability gap’ for France of €1 409m I calculate that a premium of €0.14/GWh is required per€m of liability gap.However, this liability level – €1.5bn – is not grounded on realistic evaluations of the impact of a severe nuclearaccident, as has been amply demonstrated by the events following the Fukushima disaster in March 2011. Amore relevant calculation assumes levels of financial cover such as have been reserved by TEPCO forcompensation related to the Japanese nuclear incident at Fukushima. The total issued in special bonds to cover 69the liability could reach €83bn (2005 value). These figures are supported by more recent figures from the 70Japan Centre for Economic Research, which posit a value within the range €48-168bn (2005 value).I calculate the publicly subsidised risk as the liability difference between the existing levels of private coverage,and the damage caused by the nuclear disaster at Fukushima. The steps in the calculation are:  Establish the gap between a realistic economic impact of major nuclear incident and the average insured amount for EU nuclear electricity producers  Calculate the insurance premium which would be required to cover this gap (€0.154/GWh per €m of liability gap, 2010 values) Table 8: Nuclear liability, liability gap compared with €1,500m, and generation capacity per Member State Member state Covered Liability Liability liability gap premium 71,72 73 (€m) (€m) (€m/GWh) Belgium 550 86 450 13 313 Czech Republic 296 86 704 13 35265 Nuclear Decommissioning Authority, ‘NDA Annual Report and Accounts, 2009/10’, 2010, p. 21.66 Official Journal of the European Union, ‘Opinion of the EESC on the Nuclear Illustrative Programme’, 2007.67 Leurs et al., Environmentally Harmful Support Measures in EU Members States, p. 137; Scenario ‘A’ selected forcomparison.68 Total compensation available under Brussels Convention (2004)69 The Guardian, ‘Japan Cabinet Approves Fukushima Nuclear Compensation’, 2011.70 http://www.jcer.or.jp/eng/research/policy.html71 S Carroll and A Froggatt, ‘Nuclear Third Party Insurance - the “Silent” Subsidy. State of Play and Opportunities in Europe’,2007, p. 36.72 World Nuclear Association, ‘Civil Liability for Nuclear Damage’, n.d.73 Defined as the gap between current liability levels, and the reported economic liabilities of the Fukushima disaster of€83bn 12
  14. 14. Finland 300 86 700 13 352 France 91 86 909 13 384 Germany 2 500 84 500 13 013 Hungary 183 86 817 13 370 Lithuania 50 86 950 13 390 Netherlands 522 86 478 13 318 Romania 550 86 450 13 313 Slovakia 75 86 925 13 386 Slovenia 275 86 725 13 356 Spain 275 86 725 13 356 Sweden 345 86 655 13 345 UK 275 86 725 13 356 Average 449 86 450 13 313The figures are summarised in Table 8. They demonstrate the average public subsidy value for nuclearelectricity liability in Europe is €13 329/GWh (€11 796/GWh in 2005 values).This compares rather conservatively with other studies. The Germany Renewable Energy Foundation hascalculated the cost of insuring against real liability for nuclear power in Germany as between €14 000/GWh and 74 75€230 000/GWh . The ‘Scenario B’ (upper estimate of damages) insurance cost estimated by Leurs and Witwas €50 000/GWh. h. Security of supplySecurity of supply encompasses a range of aspects, some of which (e.g. hedging, balance of payments) havealready been considered. Here I focus on the aspect of systematic risk related to the mix of electricitygenerators.There are clear benefits to the stability of an electricity system if it distributes energy from a number of differentsources. This reduces the likelihood of systematic collapse, because outages from a particular sector haveproportionally smaller impact as the number of generating sources increases. The increasing resilience of thesystem is well-described by Mean-Variance Portfolio (MVP), which is a tool widely used in financial services tomanage risk within a complex and unpredictable economic landscape. Transferring the concept to the powersector, MVP states that considering generating assets in combination rather than individually lowers the overallrisk. Efficient generating portfolios therefore maximise the expected return on investment for a given risk; or 76conversely, they minimise risk for a given return . This concept is outlined in Figure 7. 77Figure 7: Risk-cost combinations for a modelled electricity supply . The most efficient balance of cost/riskis described by the frontier of the points; in this case the point ‘H’ represents the minimum risk portfoliobetween the choices of four assets; wind, nuclear, CCGT and oil combustion.74 B Gunther et al., ‘Nuclear power plant liability in Germany’ (Bundesverband Erneuerbare Energie, 2011).75 Leurs et al., Environmentally Harmful Support Measures in EU Members States, p. 137.76 H. A Beltran, ‘Modern Portfolio Theory Applied to Electricity Resource Planning; MSc Thesis’ (University of Illinois, 2009).77 Beltran, ‘Modern Portfolio Theory Applied to Electricity Resource Planning; MSc Thesis’. 13
  15. 15. A more complete calculation was carried out in 2003 on the existing electricity generating stock in Europe(Figure 8). This indicates that system security (measured through risk) can be correlated to both cost andgenerating mix. Although there is no quantifiable output which can be derived from the literature to ascribe acost to ‘generic’ security of supply issues, this approach may suggest an avenue of further research in order toassist policy-makers with making such an informed decision. For example, I could calculate the differencebetween the current energy mix, and a set of possible European scenarios to determine the marginal impact ofa change in each electricity generating type on the portfolio return.Figure 8: EU energy portfolio, including fuel, O&M and construction period risk for new and existingcapacity. Note that a less risky portfolio can be obtained for the same return by moving horizontallyleftwards from the point of existing electricity generating infrastructure, indicated by the ‘EU 2000 gen mix’ 78boxNote that merely having a diversified energy mix does not of itself reduce overall risk unless price, technologyand correlation are properly considered; in this regard academic studies demonstrate what is suggested by 79common sense and are an extremely useful tool in analysing the system make-up.It is interesting to consider what impact the outputs of this paper, particularly the eLCOE, would have upon theMVP approach for electricity generation. The cost of each generating technology in the eLCOE is different to itsLCOE. This would presumably lead to a greater proportion of renewable energy along the minimum costfrontier, and support the argument for a greater spread of technology types (i.e. greater proportion of renewableenergy) within the energy mix. i. Virtual gridTheory suggests that the installation of decentralised generators, such as renewable energy generators willinfluence the costs to the Distribution Network Operator (DNO) in a number of ways. A full description of these 80impacts is available in the literature ; a brief summary of the main points follows.Network reinforcement, which is unnecessary at low levels of distributed generation (DG) penetration, increasesprogressively with the amount and density of DG. In rural areas this takes the place of upgraded circuits andsubstations to cope with voltage rise; urban areas require an upgrade of switchboards to provide additionalprotection.Energy losses will initially decrease as DG is introduced to the network, due to reduced power flows across it.Losses will also tend to be reduced through the use of any decentralized generator which is embedded near theend user, as losses in transmission are avoided. Solar PV is the principal contributor to this virtual grid benefit78 S Awerbuch and M Berger, ‘Applying Portfolio Theory to EU Electricity Planning and Policy-making’ (International EnergyAgency, 2003).79 M. Sunderkötter and C. Weber, ‘Valuing Fuel Diversification in Optimal Investment Policies for Electricity GenerationPortfolios’ (2009).80 Pudjianto et al., ‘Costs and Benefits of DG Connections to Grid - Studies on UK and Finnish Systems’. 14
  16. 16. because the energy is largely used onsite or close to the generator. I use European Photovoltaic Industry 81Association (EPIA) figures showing a net economic benefit of €5 000/GWh , noting that this is substantially 82lower than the suggested average benefit of €14 000/GWh made by RECABS.The increased electricity input from DG can be handled either actively or passively by a DNO. Activemanagement requires lesser capital cost but increases the system losses, as components are used to theirmaximum potential. Passive management relies on upgrading infrastructure to cope with ‘worst-case scenarios’from the point of view of large loads being produced by the DG.There is an additional economic benefit from the ‘capacity replacement value’ which mitigates the requirementto reinforce the system due to load growth, which arises from the reduced power flows across the network. Thegeneral view of the literature is that distributed generation at a small-scale (which generally refers to solarphotovoltaic, but could also apply to small-scale wind or CHP systems) has a positive economic impact; aquantitative assessment for all technologies is not possible at this stage.It is likely that electricity storage will start to play a more significant role in grid operation in the future, and willhelp define the content and concept of ‘virtual grids’. Recent work suggests that storage can obviate the need 83for regulating combustion technology, at a rate of at least 2:1 . j. Water useProviding clean drinking water requires energy, and water is needed for most energy generation types. Thisenergy-water relationship is sometimes referred to as the water-energy ‘nexus’.Water is a crucial component in Europe’s ecosystems and society. It is also a stressed resource, which is likelyto come under further strain as climate change alters the flow patterns in the future (see Figure 9). 84Figure 9: Relative change in annual river flow between reference (1961-1990) and scenario (2071-2100)In order to fully understand the interplay between water and electricity generation, I must delineate two conceptswhich are used to describe water use by the sector, namely water withdrawal and water consumption. I definewater withdrawal as the amount which is removed from a water body, regardless of whether it is returned to thewater body or totally consumed. Water consumption is defined as the net loss to the water body as a result ofelectricity generating chains; this can happen due to physical loss (such as through evaporation, inclusion incrops) or because it has become unsuitable for return due to contamination or pollution. I use the figures forconsumption by the electricity generators to calculate economic impact.Water is used in different ways by the energy sector. Thermal plants (such as biomass, coal, natural gas and81 European Photovoltaic Industry Association, Greenpeace, ‘Solar Photovoltaic Electricity Empowering the World’, 2011.82 International Energy Agency, ‘Renewable Energy Costs and Benefits for Society - RECABS’, p. 37.83 Inc Kema, ‘Research Evaluation of Wind and Solar Generation, and Storage on the California Grid’ (California EnergyCommission, 2010).84 European Environment Agency, ‘Water Resources: Quantity and Flows — SOER 2010 Thematic Assessment - ThematicAssessments — EEA’, 2010. 15
  17. 17. nuclear) require large amounts of cooling water to cool and condense the steam which drives the turbines. Theamount of water required depends on the overall plant thermal efficiency, fuel type, power plant technology andthe type of cooling. Hydropower operators use water as the ‘fuel input’ to turn the turbines. All generators usewater in their fabrication.In this study, the thermal efficiency and power plant technology is averaged amongst OECD or European plant.Regarding the type of cooling, there are two main ways for thermal plant to use cooling water; ‘once-through’and ‘recirculating’.As the name suggests, once-through systems extract water, typically from a river or the sea, and then dischargeit (with its thermal energy greatly increased) back to the water body. Leaving aside the ecological effects of thisthermal shock, there are practical implications to the electricity generation system which result from thismechanism. Any significant impediment to the water intake can cause disruption to the activity of the thermalplant. This has been amply demonstrated on a number of occasions, most recently in Germany in 2010 whenthe Unterweser nuclear power plant had to reduce its output by 60% due to high temperatures of the cooling 85water . Recirculating or closed-loop cooling systems reuse the cooling water. These technologies withdraw lesswater than once-through systems but evaporative water consumption is about 80% higher.The principal consumption of water in hydropower occurs in evaporative losses from reservoirs. This is clearly ageographically-sensitive variable, which underlines the need for the information herein to be interpreted withcaution.Figure 10: Water use of different electricity generators. Water use for the growth of biomass feedstock isexcluded, as is the water temporarily extracted from the flow of the main water body from small hydro 86,87,88,89,90,91,92(Sources: )85 www.tagesspiegel.de/politik/auch-atomkraftwerke-machen-bei-hitze-schlapp/1886228.html86 Kenneth Mulder, Nathan Hagens, and Brendan Fisher, ‘Burning Water: A Comparative Analysis of the Energy Return onWater Invested’, AMBIO, 39 (2010).87 B Sovacool, ‘Critically Weighing the Costs and Benefits of a Nuclear Renaissance’, Journal of Integrative EnvironmentalSciences, 7 (2010).88 Annette Evans, Vladimir Strezov, and Tim J. Evans, ‘Assessment of Sustainability Indicators for Renewable EnergyTechnologies’, Renewable and Sustainable Energy Reviews, 13 (2009).89 K Gerdes and C Nichols, ‘Water Requirements for Existing and Emerging Thermoelectric Plant Technologies’ (NationalEnergy Technology Laboratory, 2009).90 C Clark et al., ‘Water Use in the Development and Operation of Geothermal Power Plants’ (US Department of Energy,2010).91 Alexandre Stamford da Silva and Fernando Menezes Campello de Souza, ‘The Economics of Water Resources for theGeneration of Electricity and Other Uses’, Annals of Operations Research, 164 (2008).92 Vasilis Fthenakis and Hyung Chul Kim, ‘Life-cycle Uses of Water in U.S. Electricity Generation’, Renewable and 16
  18. 18. There are many studies on the water use of electricity generating plant. I analysed the outputs of seven differentstudies or meta-studies. There is some variation in values, which is frequently a result of methodologicalchoices or technology types; however, most studies produce values which are roughly aligned, and I take anaverage of these and discard the outliers. The results of this analysis are shown in Figure 10. Note that I do notinclude in the biomass figures the water used to grow the feedstockIn order to determine the economic impact for each generator type due to water usage, I need to establish the 93value of water. Methodologies for this calculation exist (e.g. ), but are generally complex and site-specific. Ineed to take a broader approach which uses an acceptable average for Europe as a whole. 94 3According to the OECD , the cost to the household user in Europe is approximately $1.75/m . The cost toindustry of abstracting groundwater or surfacewater is lower, and ranges from zero to something approaching 3 3 95Full Cost Recovery (FCR), such as the Netherlands (€1/m ) and Denmark (€0.55/m ) . Countries where thecharge is low generally do not include FCR which is the guiding principle behind water pricing according to theWater Framework Directive. Recent research has suggested a price increase of between 20% and 400% on 96current charges to reach FCR. In order for FCR to be met in Cyprus in 2010, predicted irrigation charges will 3 3need to be raised from €0.17/m to €0.53/m . 3With a wide spread of values for water, I select something of an average value of €0.5/m . The influence thishas on the cost of electricity is further explored in the sensitivity analysis; Table 9 summarises the costs of water 97use for each technology. Most of Europe will experience increasing water scarcity in the coming decades , so itis almost certain that the societal costs of water consumption due to electricity generation will also rise. Table 9: Cost of water consumption by electricity generation technology (2005 values) Biomass Coal Geothermal Hydro Hydro Natural Nuclear Solar Wind (small) (large) gas PVWater use 1 533 1 711 1 079 0 3 400 681 2 700 0 0 3(m /GWh)Cost (€/GWh) 678 757 477 1,504 0 301 1,195 0 05. Discussion a. Cost of electricityMy findings are clear. The economic impacts of electricity generation, above and beyond the market cost, arepositive for wind and solar PV, and negative for the other generators considered in this paper. Coal-firedelectricity emerges as the technology with the largest negative external impact (Figure 11).Sustainable Energy Reviews, 14 (2010).93 Stamford da Silva and Campello de Souza, ‘The Economics of Water Resources for the Generation of Electricity andOther Uses’.94 Organisation for Economic Cooperation and Development, ‘Pricing Water Resources and Water and Sanitation Services’,2010.95 J. Berbel, J. Calatrava, and A. Garrido, ‘Water Pricing and Irrigation: a Review of the European Experience’, IrrigationWater Pricing Policy: The Gap Between Theory and Practice. CABI, IWMI (2007).96 C. Zoumides and T. Zachariadis, ‘Irrigation Water Pricing in Southern Europe and Cyprus: The Effects of the EU CommonAgricultural Policy and the Water Framework Directive’, Cyprus Economic Policy Review, 3 (2009).97 European Environment Agency, ‘Adapting to Climate Change - SOER Thematic Assessment’, 2010, p. 15. 17
  19. 19. 60000 Water 50000 Virtual grid 40000 Risk underwriting 30000 Merit order effect Materials damage 20000 Occuaptational risk 10000 Hedging 0 Pollution-10000 CO2 Balancing costs-20000Figure 11: External costs of electricity generation, € /GWhHowever, the externalities alone do not provide the most appropriate way to consider the cost of electricity. Inorder to gain a true understanding of the picture, I need to add the externalities to the LCOE. In doing so I obtaina new ‘Extended’ Levelised Cost of Electricity (eLCOE) which is presented in Figure 12. eLCOE in 2010 (€/GWh, 2005 prices) 450,000 External 400,000 350,000 impact 300,000 250,000 200,000 150,000 100,000 50,000 0 -50,000Figure 12: Extended levelised cost of electricity (eLCOE), i.e. the LCOE plus externalities. Negative externalitiesindicate that there is an economic benefit above and beyond the production of electricity itself. The error barsindicate the range of costs arising from a sensitivity analysis.The result of including a wide range of externalities into the costs of electricity production demonstrate that‘business as usual’ burdens human health and the environment with large costs, and that a transition to a low-carbon energy system will have multiple benefits beyond simple CO2 reduction.I exclude from this analysis those factors which are already incorporated into pricing mechanisms, such as costsfor forecast error in variable generators. In other words, I assume that these are already incorporated into theLCOE. I am also unable to calculate the impact for a number of external factors, including ecosystem impacts,second-order climate change impacts of combustion such as increased sanitation and parasite-related healthproblems, desertification, deforestation etc; and the health impacts from a number of combustion by-productssuch as PCBs, Cu, Se and Zn.Using existing references to provide an average value for LCOE in 2020, I can also generate an eLCOE for2020 (. I assume that external impacts remain the same with the exception of pollution costs, which are reducedby 50% due to improved technology and efficiency. The significant change is the overall cost of solar PV, andthe improved position of geothermal with respect to conventional generation. 18
  20. 20. eLCOE in 2020 (€/GWh, 2005 prices) 250,000 External impact LCOE 200,000 150,000 100,000 50,000 0 -50,000Figure 13: Extended levelised cost of electricity (eLCOE), i.e. the LCOE plus externalities. Negative externalitiesindicate that there is an economic benefit above and beyond the production of electricity itself. The error barsindicate the range of costs arising from a sensitivity analysis. b. Employment 98The creation of new jobs is increasingly being seen as a positive ‘externality’ of activity in different sectors , andthere is considerable enthusiasm in the renewable energy sector for being able to demonstrate a highemployment factor for the sector.The value of employment to an economy is defined through the internationally agreed System of National 99Accounts . Within this framework, individuals supply labour that generates compensation. This compensationis a monetary unit which is the component of the ‘generation of income’ account directly attributable toemployment. Simply put, the income derived from a paid activity can be considered a direct economic benefitequivalent to the value of the income. However, there is sufficient uncertainty in the accuracy of the estimates ofjob creation to lead me to exclude this component from the overall value, and instead to use it as an externalindicator of additional value.To determine this value, I need to calculate the number of jobs created through the construction and operationof electricity generating plant (levelised over its lifetime), and the value of the employment to the individual.I use the outputs of a meta-study aggregating many different individual assessments on employment per 100GWh over the lifetime of the generating technology. This paper calculates direct and indirect jobs only;induced employment (which is that created by economic activity carried out by direct and indirect employment)is not included. Jobs within the offshore wind and large hydro sectors are also not provided. 101I assume a sector average salary for renewable energy employees of €58 601 . Although the salary for therenewable energy sector is US-based (rather than European), it represents a relatively up-to-date figure, and isprobably broadly representative of the European figures. I also assume that this salary is representative ofwages in the ‘conventional’ energy sector.I multiply the average job-years/GWh by the average energy sector salary to obtain an employment value for98 C. Tourkolias and S. Mirasgedis, ‘Quantification and Monetization of Employment Benefits Associated with RenewableEnergy Technologies in Greece’, Renewable and Sustainable Energy Reviews, 15 (2011).99 EC, IMF, OECD, UN, WB, ‘System of National Accounts 2008’ (EC, IMF, OECD, UN, WB, 2009), p. 22,33.100 Adapted from Max Wei, Shana Patadia, and Daniel M. Kammen, ‘Putting Renewables and Energy Efficiency to Work:How Many Jobs Can the Clean Energy Industry Generate in the US?’, Energy Policy, 38 (2010).101 Simply Hired, ‘Renewable Energy Salaries’, 2011. 19
  21. 21. each electricity generating technology, which is then levelised by the electricity output (Table 10).Table 10: Employment value/GWhof different electricity generatorsEnergy Average Annualtechnology job years/ value (€) GWhBiomass 0.21 12 306Coal 0.11 6 446Geothermal 0.25 14 650Natural gas 0.11 6 446Nuclear 0.14 8 204Small hydro 0.27 15 822Solar PV 0.87 50 982Wind 0.17 9 962This demonstrates that labour-intensive operations (particular solar PV) generate significant employmentbenefits. However, this is a rather simplistic way of considering the issue. More relevant to most policymakerswill be the economic benefit from employment, levelised by both the electricity output and the cost of thatelectricity.This additional step is calculated by dividing the economic benefit for each technology by the LCOE. Thisprovides me with an ‘Electricity Employment Factor’ (EEF) presented in Figure 14. This factor is dimensionless,as I am dividing a value of employment in €/GWh by the LCOE, which has the same units.Figure 14: EEF ( levelised by electricity generation and LCOE)However, this picture is, once again, too simplistic, because the use of LCOE does not include the externalitieswhich accrue from using each electricity generating technology. The most relevant metric is therefore theemployment value, divided by eLCOE (Figure 15). The main difference between the LCOE and eLCOE figuresfor employment is the improved position of biomass compared with nuclear. 20
  22. 22. Figure 15:EEF (levelised by electricity generation and ELCOE)Questions have been raised about whether net jobs are created by the renewable energy sector, or whetherthey are merely displaced from other sectors. The results of four studies in Germany are clear; that the sector 102creates significant additional employment above and beyond any displacement effect . My calculationsvalidate this result. All these developments take place within the same sector, and so the comparisondemonstrated by Figure 15 is valid; in other words, employment on a hydropower project creates jobs additionalto those that would have been created from producing the same amount of electricity in a coal-fired generator.The marginal benefit is the difference between the EEF of the different generating types.I can therefore conclude that when externalities are taken into account, renewable electricity generatorsdemonstrate a very strong economic benefit related to employment, compared with conventional generators.6. AcknowledgementsI would like to thank Rina Bohle Zeller of Vestas for her comprehensive knowledge on the water use of differentelectricity generating technologies. I would also like to thank Alfredo Sanchez Vicente, Arthur Girling, BittenSerena and Cinzia Pastorello for their personal support during the preparation of the paper.This work was carried out during my period of employment at the European Environment Agency, between 2010and 2012.102 Federal Environment Agency, Germany, ‘Report on the Environmental Economy 2009’, 2009, p. 32. 21